"background": "Public health surveillance systems are critical for disease control, yet their methodological reliability in longitudinal application within complex, resource-variable settings is inadequately characterised. This is particularly pertinent in sub-Saharan Africa, where system performance directly impacts health outcomes. ", "purpose and objectives": "This study aimed to conduct a longitudinal methodological evaluation of a national public health surveillance system. Its primary objective was to quantify system reliability over time using a multilevel modelling framework, identifying facility- and district-level determinants of reporting consistency. ", "methodology": "We employed a longitudinal cohort design, tracking reporting entities over a multi-decade period. System reliability was operationalised as a composite score of timeliness, completeness, and accuracy. A three-level random intercepts model was fitted: Y{ijt = \0 + \ Xijt + uj + vk +, where i, j, and k index reports, facilities, and districts, respectively. Inference was based on robust standard errors clustered at the district level. ", "findings": "Analysis indicates a significant positive longitudinal trend in overall reliability, with a 22% improvement in the mean composite score. However, substantial heterogeneity persisted; facility-level resource allocation was a stronger predictor (β = 0. 31, 95% CI: 0. 24, 0. 38) than district-level governance indicators. The intra-class correlation suggested 40% of the variance in reliability scores was attributable to district-level factors. ", "conclusion": "The surveillance system demonstrated measurable improvement in methodological reliability, yet inequities rooted in structural and resource determinants remain entrenched. System strengthening must address multilevel influences to ensure consistent performance. ", "recommendations": "Resource allocation should target facility-level capacity as a priority. Monitoring frameworks must integrate multilevel reliability metrics. Future evaluations should employ similar longitudinal, hierarchical models to disentangle systemic drivers of performance. ", "key words": "public health surveillance, system reliability, longitudinal evaluation, multilevel
Zyl et al. (Wed,) studied this question.